estimation of path attenuation and site …...seismic activity of this region can be understood...
TRANSCRIPT
Estimation of path attenuation and site characteristics in the
north-west Himalaya and its adjoining area within Indian
Territory using generalized inversion method
Harinarayan, NH1, Abhishek Kumar
*2
1, 2 Department of Civil Engineering, Indian Institute of Technology Guwahati, Assam, India.
Corresponding to: Kumar Abhishek ([email protected]/ [email protected])
Abstract. Present work focuses on the determination of path attenuation and site characteristics of earthquake 1
recording stations, located in the north-west Himalaya and its adjoining region, within India. The work is done 2
using two-step generalized inversion technique. In the first step of inversion, non-parametric attenuation curves 3
are developed. ( ) ( ) as S wave quality factor within 105km, is obtained indicating that 4
the region is possibly heterogeneous as well as seismically active. In addition, presence of a kink is observed at 5
around 105km hypocentral distance while correlating path attenuation with the hypocentral distance, indicating 6
the presence of Moho discontinuity in the region. In the second step of inversion, site amplification curves are 7
developed separately from the attenuation corrected data for horizontal and vertical components of the 8
accelerograms. The amplification function and predominant frequency of each recording station based on 9
generalized inversion method is estimated and compared with that obtained from horizontal to vertical spectral 10
ratio (of S wave portion of the accelerograms) method. Path attenuation and site characteristics obtained in the 11
present study are very essential for developing regional ground motion model and for the seismic hazard 12
assessment of the above study area. 13
Keywords: Seismicity, Northwest Himalaya, path attenuation, site characteristics, Generalized Inversion, HVSR 14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
1 Introduction 32
The Himalayan arc extending approximately 2500km between Kashmir and Arunachal Pradesh is one among 33
the seismically most active regions across the globe. Seismic activity of this region can be understood based on 34
induced damages witnessed primarily during 4 great earthquakes (EQs) including 1897 Shillong EQ, 1905 35
Kangra EQ, 1950 Assam EQ and 1934 Bihar-Nepal EQ, occurred in the last 120 years. Based on seismic 36
activity, the entire Himalayan belt can be subdivided into three distinct segments namely the western, the central 37
and the eastern Himalayas (Philip, 2014). The region of the north-west Himalaya and its foothills, within India 38
come under seismic zone IV and V as per IS 1893: 2016, indicating regions of high to very high seismicity. 39
Intensity of ground shaking during an EQ, at a particular site is a collective effect of source, path and 40
site parameters. Source parameters include magnitude, fault mechanism, stress drop and rupture process. On the 41
other hand, path parameters include geometric attenuation and loss of seismic energy due to the anelasticity of 42
the earth and scattering of elastic waves in heterogeneous media. Similarly, site characteristics include 43
modification of amplitude, frequency content and duration of the incoming seismic wave by subsurface medium 44
as reaches the surface. Determinations of aforementioned EQ parameters are important for the development of 45
region specific ground motion models, which can further be used for region/site specific seismic hazard 46
assessment (Baro et al., 2018). Above parameters can be estimated from EQ records based on some spectral 47
modelling or inversion approach like generalized inversion method (Andrews, 1986; Castro et al., 1990; Oth et 48
al., 2009 etc.). 49
To understand the on-going seismicity of various regions within India, the Government of India has 50
installed number of EQ recording stations. EQ records from these stations since 2004 are maintained by 51
PESMOS (Program for Excellence in Strong Motion Studies), which is currently one of the most significant 52
resource of ground motion records in India. At present, PESMOS manages EQ records from 300 recording 53
stations which are distributed in the northern and northeaster parts of India as well as in the Andaman and 54
Nicobar Islands (Kumar et al. 2012). It must be highlighted here that PESMOS database is lacking in terms of 55
accurate information about subsurface for majority of recording stations (Harinarayan and Kumar, 2018). Site 56
class given by PESMOS is based on physical description of surface materials, local geology following 57
Seismotectonic Atlas of India (GIS 2000) and Geological Maps of Indian (GSI, 1998) and not based on actual 58
field investigation (Kumar et al., 2012). In the absence of accurate information of local soil, utilizing EQ records 59
from PESMOS database for seismic studies is a major challenge. Geophysical subsurface exploration studies on 60
some of the recording stations in northwest Himalaya reported by Pandey et al., (2016a; b) had highlighted the 61
flaws in site class given by PESMOS. There are recording stations classified to be on rock site but were found to 62
be on soil sites by Pandey et al., (2016a; b). Harinarayan and Kumar (2018) also reported problems in the 63
subsurface information given by PESMOS and attempted site classification of PESMOS recording stations 64
In the present study, EQ records from the region of the north-west Himalaya and its foothills within 65
India, as obtained from PESMOS database, are analysed for estimating path attenuation and site characteristics 66
separately, using a two-step generalized inversion of the S-wave Fourier spectra (hereafter referred to as GINV). 67
In the first step, attenuation curves are developed using a non-parametric inversion approach (similar to Castro 68
et al., 1990 and Oth et al., 2008). In the conventional generalized inversion method (Andrews, 1986; Hartzell, 69
1992), the second step of inversion calculates both site and source spectra, by inverting the S-wave (or Coda 70
wave) spectra, corrected for the path parameter. This method however requires one or more reference sites 71
(usually rock sites) in-order to remove the trade-off between the source and site parameters (Andrews 1986). In 72
the absence of subsoil information for majority of recording stations managed by PESMOS as highlighted 73
above, identifying reference site is not possible. For this reason, only the site parameters are evaluated in the 74
second step of inversion, using a non-reference generalized inversion approach (similar to the work by Joshi et 75
al., 2010; Harinarayan and Kumar 2017b). Further, Obtained site terms are compared with the one calculated 76
from horizontal to vertical spectral ratios (HVSR) from the same S-wave window as used in the GINV. 77
This study is one of its kind, which systematically evaluates path and site parameters using a larger and 78
regional database. Existing studies on attenuation characteristics determination for the northwest Himalaya used 79
few EQ records that too from limited recording stations. Joshi (2006) estimated frequency independent S wave 80
quality factor ( ) for the Garhwal Himalayas using 1991 Uttarkashi EQ and 1999 Chamoli EQ ground motion 81
records from 8 recording stations. In another study, Singh et al., (2012) estimated frequency depended for the 82
Kumaun Himalayas using 23 EQ events, from 9 recording stations by applying the extended coda-normalization 83
method. Similarly, Negi et al., (2015), Banerjee and Kumar, (2017) and Tripathi et al., (2015) estimated for 84
the Garhwal Himalayas. The aforementioned studies did not highlight the attenuation characteristics of the 85
entire north-west Himalaya region. 86
Similar to path attenuation studies, very few studies on the determination of site characteristics from 87
EQ records also exist for this region. Nath et al., (2002) computed site terms using the aftershocks of the 1999 88
Chamoli EQ, obtained from 5 recording stations located in the Uttarakhand region. Similarly, Sharma et al., 89
(2014) estimated site parameters for the Garhwal region of Uttarakhand using EQ records in context of 90
generalized inversion and HVSR. In another work, Harinarayan and Kumar (2017) reported a comparative study 91
on site characteristics computed using EQ records from Tarai region of Uttarakhand using multiple analytical 92
approaches. In another recent work, Harinarayan and Kumar (2018) computed site parameters for recording 93
stations in the northwest Himalayas in terms of predominant frequency (fpeak) alone using HVSR method. 94
2 Study Area 95
Present study area includes states of Himachal Pradesh, Uttarakhand, Punjab, Haryana and national capital city, 96
New Delhi, covering an area between 28º N to 34º N latitude and 75.8º E to 80.5º E longitude. According to 97
2011 Census, the region has a population of 96 million. From seismicity point of view, major Seismotectonic 98
features of the present study are characterized by three north-dipping thrust systems such as the Central thrust 99
(MCT), the Main Boundary thrust (MBT) and the Himalayan frontal thrust (HFT) (Valdiya, 1981). Other 100
tectonic features includes, the Jhelum Balakot fault, the Drang thrust, the Lesser Himalayan Crystalline Nappes, 101
the Jammu thrust, the Vaikrita thrust, the Karakoram fault, the Jwala Mukhi thrust, and the Ramgarh thrust. 102
Both the MCT and the MBT lie parallel to each other within western Himalayan region and were produced 103
during the Cenozonic shortening (Malik and Nakata 2003). The HFT is the youngest active thrust separating the 104
Himalaya region and the Indo-Gangetic alluvial plain (Kumar et al. 2009). The HFT, the MBT and the MCT 105
have generated major EQs in this region (Philip et al. 2014). 106
Two of the most damage inducing EQs, in the last 120 years, in the west Himalayan regions include 107
1905 Kangra-Himachal Pradesh EQ (Ms=7.8) (Ambraseys and Douglas 2004) and 2005 Muzzafarbad-Kashmir 108
EQ (Mw=7.6) (Avouac et al. 2006). Both of these EQs caused severe loss to life and property. Recent EQs of 109
1991 Uttarkashi (Mw=6.8) and 1999 Chamoli (Mw=6.6) had occurred on the MCT zone (Harbindu et al. 2014). 110
The 1999 Chamoli EQ caused a huge landslide in Gopeshwar situated less than 2km northwest of Chamoli city 111
(Sarkar et al. 2001). This EQ also caused shaking in Chandigarh and Delhi, located far away from the epicentre 112
(Mundepi et al., 2010). 113
3 Database 114
Ground motion records used in this study consists of three components accelerograms obtained from PESMOS 115
database available at http://www.pesmos.in/. The instrumentation used for recording EQs consists of internal 116
AC-63 GeoSIG triaxial force balanced accelerometers and GSR-18 GeoSIG 18 bit digitizers with external GPS 117
(Kumar et al., 2012). Further, ground motion recordings are done in trigger mode during each EQ with a 118
sampling rate of 200 per second. 119
For the present analysis, ground motion records of EQs happened between 2004 and 2017, available on 120
PESMOS are used. For estimating site characteristics, 341 records from 86 EQs, with magnitudes ranging from 121
Mw=2.3 to 5.8, having focal depths ranging from 2 to 80km are used. Further, these records are corresponding 122
to 101 recording stations, located in the hypocentral distance ranging from 9 to 355km. Coordinates of each of 123
the recording station, used in this work are listed in Table 1, columns 2 and 3. Further, details of EQs used for 124
estimating site characteristics are summarized in Table 2. 125
For estimating path attenuation however, only those EQs recorded at atleast two recording stations 126
among which at least one recording station is located within hypocentral distance equal to or less than the 127
reference distance (reference distance is discussed under section ‗Spectral attenuation with distance) are 128
considered. Out of 341 EQ records used for estimating site parameters, only 207 EQ records satisfies the above 129
mentioned reference distance criteria. Final database for estimating path attenuation consists of 207 records 130
from 32 EQs, recorded at 69 recording stations, with magnitude (Mw) in the range of 3.1 to 5.5, focal depths 131
from 3 to 55km and hypocentral distance from 9 to 200km. Table 3 summarizes the details of the dataset used 132
for estimating path attenuation. In addition, Figure 1 shows the source-to-recording station distance of the data 133
set used in the present study. 134
3.1 Data processing 135
Signal to pre-event noise (all of equal window length) ratio (SNR) for all the records are computed and records 136
with SNR≥5 (similar to Ameri et al., 2011) are considered for further analyses. All the EQ records are corrected 137
for baseline correction following a 5% cosine taper and a band-pass filtering, between the frequency range of 138
0.25Hz and 15Hz, using a Butterworth filter. Further, time windows starting about 0.5s before the onset of the S 139
wave and ending when 90% of the total seismic energy of the EQ record is reached, are separated and tapered 140
with a 5% cosine window (Ameri et al., 2011; Bindi et al., 2009). Typical lengths of the time windows for the 141
present analysis vary from 4 to 15s. Further, for some of the records, where the window lengths obtained to be 142
longer than 15s, are fixed to 15s in order to minimize coda wave energy in the analysing time window (Oth et 143
al., 2008). Later, based on the extracted windows, the Fourier amplitude spectra is calculated for each EQ 144
record, smoothened by applying the Konno and Ohmachi (1999) algorithm, with the smoothing parameter ―b‖ = 145
20. 146
For further analyses, path and site parameters are estimated using above processed EQ records, based on two 147
separate inversion procedures as discussed separately in the following sections. 148
4 Path attenuation 149
In the first step of inversion, path attenuation curves are developed by eliminating the effect of site parameter, 150
thereby retaining only the source and path attenuation characteristics. All EQ records, irrespective of whether 151
located on soil or rock sites can be utilized for inversion. This way, present method is very much suitable for 152
PESMOS database where accurate subsurface information of recording stations is not available. The horizontal 153
portion of the accelerograms (obtained by the root mean square average of the east-west and north-south 154
components) is considered for developing path attenuation curves. Detailed discussions on the method can be 155
found in following sub-section. 156
4.1 Methodology 157
Following Castro et al., (1990), observed spectral amplitude (acceleration) ( ), of EQ , at recording 158
station , and frequency f can be modelled linearly as: 159
( ) ( ) ( ) (1) 160
Here, ( ) is a scalar, which is governed by the magnitude of the EQ (one value for each EQ). ( ) is the 161
attenuation function and is independent of the magnitude of EQ. Here, ( ) incorporates both geometric 162
spreading and anelastic attenuation variation with the hypocentral distance. It has to be mentioned here that 163
( ) in Eq. (1) is not limited to a particular functional form, instead, is assumed to decay smoothly with 164
hypocentral distance ( ) and thus take the value of unity at a reference distance ( ), as given in Eq. 2 (Castro 165
et al., 1990; 1996; 2003). 166
( ) (2) 167
Model given by Eq. (1) has no factor representing site effect and is contained in both ( ) 168
and ( ). Any rapid undulations in ( ) are due to the absorbed site effects (Oth et al., 2008). Two 169
weighing factors, and are incorporated in the Eq. (1) following Castro et al., (1990). is used to 170
smoothen the attenuation term with distance curve by supressing the undulations and thereby removing any 171
absorbed site effects from ( ). is used to impose ( ) constraint, as mentioned earlier. The 172
value of and here is chosen reasonably such that the site effects are supressed but the change in the 173
attenuation characteristics with distance can be observed (Oth et al., 2008). Solution to Eq. (1) in the matrix 174
form, after incorporating weighing factors, is obtained using singular value decomposition method, (discussed in 175
detail in Appendix A). 176
4.2 Spectral attenuation with distance 177
Figure 2 shows the number of EQ records for various hypocentral distance range considered. It can be observed 178
from Figure 2 that there are very less EQ records available beyond hypocentral distance of 115km. For this 179
reason, EQ records with hypocentral distance up to 115km are only considered for the determination of path 180
attenuation. The constraint ( ) is applied at =15km, irrespective of the frequency. The hypocentral 181
distance range from 15 to 115km is divided into 10 bins, each bin having 10km width. Further, attenuation 182
curves are computed for each of the selected 17 frequencies from 1Hz to 15Hz (see Table 4, column 1). 183
Variation of attenuation curves with hypocentral distance, obtained in the present study, for the selected 184
frequencies is depicted in Figure 3. Based on Figure 3, a general trend in which attenuation curves exhibit decay 185
with hypocentral distance up to 105km can be observed. Beyond 105km a kink is observed as seen in Figure 3. 186
This kink in the attenuation curves beyond 105km is very distinct and clear at lower frequencies (<5.5 Hz). 187
Bindi et al., (2004) and Oth et al., (2011) reported a similar trend in the attenuation curves for the Umbria 188
Marche and Japan regions respectively. Oth et al., (2011) attributed this behaviour to the combined effect of 189
reflected or refracted wave arrivals from the Moho in Japan. Presence of Moho in the North-west Himalaya was 190
reported by Saikia et al., (2016) based on Teleseismic receiver function analysis. Referring to Oth et al., (2011) 191
work, presence of kink beyond 105km in attenuation curves, as obtained in this study may also be due reflected 192
or refracted waves from the Moho. Further, detailed study in this direction can be done in the future and is 193
beyond the presence scope of the work. Observing the attenuation curves at different frequencies as given in 194
Figure 3, it can concluded that attenuation curves at higher frequencies (>5.5Hz) decay more rapidly compared 195
to lower frequencies for the present study region. This observation is consistent with the findings by Castro et 196
al., (2003) for Guadeloupe (France) and Oth et al., (2011) for Japan. 197
Further, for the kink observed at 105km, in case of lower frequencies, its sharpness reduces with 198
increasing frequency, as can be observed in Figure 3. At frequencies greater than 10Hz, the kink at 105km 199
smoothen and the attenuation curves beyond 105km for frequencies greater than 10 Hz becomes flat as observed 200
in Figure 3. This change in the characteristics of the kink at higher frequency indicates that the arrival of waves 201
from the Moho also get attenuated more at higher frequencies in comparison to lower frequencies. 202
4.3 Quality factor ( ) estimation 203
In order to estimate , inversion is repeated, however only considering records within hypocentral distance in 204
the range 15km to 105km, where a monotonic decrease in attenuation curves with hypocentral distance is 205
observed (see Figure 3). The attenuation curves are modelled in terms of geometric spreading ( ) and 206
quality factor ( ) in accordance with Castro et al., (1996) as; 207
( ) ( ) [
] (3) 208
Where, is the frequency and is the mean shear wave velocity in the crustal medium taken as 3.5km/s as per 209
Mukhopadhyay and Kayal, (2003). Further, ( ) is considered as in accordance with Banerjee and 210
Kumar (2015) for this region. For each frequency considered in this study (see Table 4), Eq. (3) is linearized by 211
taking logarithm and corrected for the effect of ( ) as given in Eq. (4). 212
( ) ( )
(4) 213
Ascribed to Castro et al., (2003), Eq. 4 is written in the form; 214
( ) (5) 215
Where ( ) and are given as; 216
( ) ( ) ( ) (6) 217
(7) 218
Where, m in Eq. 5 represents the slope between ( ) and based on a linear least-square fit, obtained for each 219
of the selected frequencies. Further, the values are estimated for the selected frequencies by substituting the 220
value of computed using Eq. (7). Columns 2 and 3, Table 4 list the value of and with frequency ( ) 221
respectively. In order to build the frequency dependent relationship ( ), the value of is fitted as a 222
function of frequency using a power law. In the above expression, is the frequency dependent coefficient, 223
which is approximately equal to 1 and varies on the basis of the heterogeneity of the medium (Aki 1980). 224
Variation of against frequency as illustrated in Figure 4 gives frequency dependent for the north-west 225
Himalayas as; 226
( ) ( ) (8) 227
Values of and (in the expression ) are attributed to the level of tectonic activity and degree of 228
heterogeneity respectively, present in the region. Aki (1980) concluded higher values of for tectonically active 229
regions in comparison to that of stable regions. Similarly, low value of (<200) is an indication of larger 230
degree of heterogeneities in the medium (Joshi 2006). The values of (=0.94) and (=105), obtained in this 231
study indicate that the present study region is tectonically active, characterized by higher degree of 232
heterogeneities, in accordance with Aki (1980) and Joshi (2006). 233
4.4 Comparison with regional and global attenuation characteristics 234
As discussed earlier, numerous studies exist where path attenuation of different parts of the present study area 235
were attempted in the past. Comparison of present results with those obtained by the previous researchers for the 236
northwest Himalaya and Delhi region are attempted as shown in Figure 5. It can be seen from Figure 5 that the 237
attenuation curve obtained in the present study falls in between existing attenuation curves for the different parts 238
of the north-west Himalaya as given in the literature, [Kinnaur, (Kumar et al., 2009), Kumoan (Mukhopadhyay 239
et al., 2010), Garhwal regions (Negi et al., 2015) and Delhi (Sharma et al., 2015)]. It has to be highlighted here 240
that the database for the present study also includes EQ records from Kinnaur, Kumoan, Garhwal regions of the 241
north-west Himalayas as well as from regions in and around Delhi. For this reason, the value of and 242
obtained in the present study reflects an average attenuation characteristics of regions encompassing north-west 243
Himalaya up to Delhi region but within Indian boundary. 244
Furthermore, the attenuation results obtained in this study is compared with some typical results 245
obtained globally in terms of attenuation characteristics and tectonic setting as shown in Figure 6. Literature 246
suggests low values of for tectonically active regions [e.g. Kato Japan region (Yoshimoto et al., 1993); East 247
central Iran (Mahood et al., 2009); Egypt (Abdel 2009); and Umbria–Marche region (Lorenzo et al., 2013)]. 248
Similarly, relatively high values of were reported for tectonically stable areas [e.g. Baltic Shield (Kvamme 249
and Havskov 1989); Central South Korea (Kim et al., 2004) and South Eastern Korea (Chung and Sato 2001)]. 250
Attenuation values obtained in the present study show good agreement with other studies having lower 251
across the globe. Further, attenuation curves for the present region is also found closer to regions of high 252
seismicity like Umbria–Marche and Eastern Iran as can be observed from Figure 6. 253
5 Site Effects 254
After estimation of path parameter as discussed in the previous section, site characteristics of the recording 255
stations are determined in the second step of inversion. In addition to GINV, site components are also estimated 256
using HVSR method. Detailed discussion on GINV and HVSR is given in the subsequent sections. 257
5.1 GINV 258
GINV was developed by Andrews (1986) by improvising spectral ratio method. Since then, various forms of 259
this technique have been developed and used for estimating the seismic site characteristics by various 260
researchers (Castro et al., 1990; Boatwright et al., 1991; Oth et al., 2008 etc.). The methodology used for 261
estimating site characteristics in the present study is discussed here. 262
As per Iwata and Irikura (1988), the Fourier amplitude (acceleration) spectrum (FAS) of the EQ 263
recorded, at the recording station, ( ) can be represented in the frequency domain as the product of 264
source term ( ( ) ), path attenuation ( ( ) ) and site term ( ( ) ) as shown below; 265
( ) ( ) ( ) ( ) (9) 266
Further, the path attenuation term can be removed from the spectral content of the record following Andrews 267
(1986) as; 268
( ) ( )
( ) ( ) ( ) (10) 269
The value of ( ) can be estimated using Eq. (3) and by considering as per Eq. (8), obtained in the earlier 270
section. Further, Eq. (10) can be linearized, by taking natural logarithms on both sides as per Andrews (1986) 271
giving; 272
( ) ( ) ( ) (11) 273
Considering: ( ), ( ) ( ) and ( ) j, Eq. (11) in the matrix form can be 274
written in accordance with Joshi et al., (2010) and following the notations of Menke (1989) as; 275
← 1st event → ← nth event → ← Site effect→
1 2 … m
1 2 … m 1 2 … m
1 0 … 0 …... 0 0
0 1 0 … 0
s1(f1)
d1(f1)
0 1
0 …... 0 0
0 0 1 … 0
:
:
:
:
: :
: : :
:
:
:
:
:
: :
: : :
:
:
:
0 0 0 1
0 0 … 0 0 0 … 1
s1(fn)
d1(fm)
sn(f1) =
:
For nth earthquake
sn(fn)
dn(fm)
0 0 … 0 ….. 1
… 0 1 0 … 0
g(f1)
dn(fm)
0 0 … 0 ….. 0 1 … 0 0 1 … 0
g(f2)
:
: :
:
: :
: : :
:
:
:
: :
:
: :
: : :
:
:
:
0 0 … 0 … 0 0 … 1 0 0 … 1
g(fm)
dn(fm)
(12) 276
The matrix form in Eq. (12) represents a purely under determinate system since there are ( ) 277
parameters for ‗ data (here is the number of sample frequency and is the number of EQs recorded at 278
a particular recording station). In here, Eq. (12) is solved using Moore- Penrose matrix inversion procedure 279
(minimum norm inversion) given by Penrose, (1955) to determine ( ) at each of the recording station. 280
Based on the above discussed methodology, inversions are performed for east-west, north-south and 281
vertical components of EQ records separately to obtain the amplification curves in the frequency range of 282
0.25Hz to 15Hz, for each of the three components. For further calculation, the horizontal component is obtained 283
as the geometric mean of east-west and north-south components. 284
5.2 HVSR 285
HVSR method is an extension of Nakamura (1989) technique, which is widely to assess the subsoil 286
characteristics using recorded ambient noises. Nakamura (1989) technique is based on the assumption that the 287
soil amplification effects are retained only in the horizontal component whereas the source and the path effects 288
are maintained both in vertical as well as horizontal components of ground motion. Hence, the ratio of 289
horizontal and vertical components gives an estimate of site amplification. Lermo and Chavez-Garccia (1993) 290
extended Nakamura (1989) technique to S wave part of the accelerograms and studied the theoretical basis of 291
the technique by numerical modelling oSV waves. Later, HVSR method was applied to EQ recordings 292
worldwide (Luzi et al., 2011; Yaghmaei-Sabegh and Tsang 2011; D‘Alessandro et al., 2012; Harinarayan and 293
Kumar 2017a, b, 2018 etc.) to obtain the site characteristics. 294
Comparative studies between HVSR and other methods of evaluating site parameters reported by Field and 295
Jacob (1995), Parolai et al., (2004), Shoji and Kamiyama (2002) Harinarayan and Kumar (2017b) etc. show that, 296
HVSR can provide good and reliable estimate of predominant frequency. However, the above literatures also 297
point out discrepancies in amplification levels obtained from HVSR with other methods. In order to compare the 298
site amplification functions obtained from HVSR and GINV methods, HVSR for each station is computed 299
considering the same S wave window as used in the GINV method. HVSR for each recording station is 300
determined using the following steps; 301
1. Calculate the response spectra considering 5% damping, for all the three components (north-south, east-302
west and vertical) of ground motion records. 303
2. Obtain the geometric mean of the two horizontal response spectra components (H) using Eq. (13) given 304
below; 305
( ) (13) 306
3. Calculate the ratio of to ( ). 307
Where, HEW and HNS are the response spectrum of the horizontal east-west and north-south components 308
respectively and V is the response spectrum corresponding to vertical component of ground motion. Then, the 309
HVSR at each of the recording station can be estimated as; 310
( ) ∑
(14) 311
Here, is the number of events recorded at recording station ― ‖ and ( ) indicates the average HVSR 312
value for a particular recording station ― ‖. The fpeak is the value of frequency corresponding to a maximum 313
value of (denoted by Apeak) at the recording station ― ‖. 314
5.3 Site Parameters 315
Site amplitude (SA) curves are developed using GINV for the horizontal (GINV H) and the vertical components 316
(GINV V). Figure 7 shows typical amplification curves obtained for GINV H (indicated by dashed lines) and 317
GINV V (indicated by firm lines) for typical 6 recording stations. In general, obtained amplification values for 318
GINV H is greater than GINV V for all frequencies. A typical observation (from Figure 7) for both GINV H and 319
GINV V is that the high level of amplification is observed at high frequencies. For several recording stations, 320
clear and distinct peak in the amplification curve can be observed (e.g. JAMI, BAR, GHA and SND) from 321
Figure 7. Moreover, the overall peaks of the GINV V component are usually at higher frequencies than for the 322
GINV H component. Further, in case of few recording stations, the shift in frequency is relatively close to a 323
factor √ (eg. BAR and GHA). Next, Site amplification factor (SAF) is estimated based on the GINV results 324
(denoted by GINV H/V) as the ratio of GINV H to GINV V. The value of frequency corresponding to the 325
maximum value of SAF (denoted as Apeak) is fpeak. GINV H/V curves are compared with those estimated using 326
HVSR method for a total of 101 recording stations. Figure 8 shows the comparison of the HVSR (indicated by 327
dashed lines) and GINV H/V (indicated by firm line) for typical 9 recording stations from above analyses. A 328
general observation made from Figure 8 is that both HVSR and GINV H/V show similar SAF patterns for all 329
recoding stations in the selected frequency range. Overall value of fpeak obtained exhibit 1:1 matching between 330
the two methods. However, there is trend of difference in terms of Apeak values. Apeak values obtained using 331
HVSR are found higher compared to those obtained using GINV H/V curves. This observation is also reported 332
in other regions (Sharma et al., 2014; Field and Jacob, 1995). The values of fpeak obtained using GINV H/V and 333
HVSR are tabulated in Column 5 and 7, Table 1 respectively. Similarly, the values of Apeak obtained using 334
GINV H/V and HVSR are tabulated in Column 6 and 8, Table 1 respectively. The maximum value of fpeak of 335
15Hz is observed for the recording station GGI with a value of Apeak of 5.3 based on GINV H/V. The maximum 336
value of Apeak of 12.2, based on GINV H/V is observed for ADIB recording station at 6.3Hz. The range of Apeak 337
based on HVSR varies between 1.7 and 19.4, while based on GINV H/V, the range of Apeak varies between 1.5 338
and 12.7. Similarly, of the fpeak based on HVSR varies between 0.4Hz and 10Hz, while based on GINV H/V, the 339
fpeak varies between 0.5Hz and 15Hz. 340
Further, based on the value of fpeak obtained above, using GINV, the recording stations are classified as 341
either rock sites or soil sites. In general, criteria based on average shear wave velocity over 30m ( ) is used 342
for site classification. A site can also be classified based on fpeak values. Such an approach was used by 343
Harinarayan and Kumar (2018) to classify the recording stations in the north-west Himalaya based on fpeak 344
obtained using HVSR method, where stations having fpeak less than 6.35Hz were classified as soil sites and 345
stations having fpeak greater than 6.35Hz were classified as rock sites. The range of fpeak values reported by 346
Harinarayan and Kumar (2018) for soil and rock sites were calculated based on the range of based on 347
NEHRP site classification scheme (BSSC, 2003) in accordance with the Eq. (15) (Kramer, 1996), correlating 348
to soil depth (denoted by and taken as 30m) and shear wave velocity ( ). 349
⁄ (15) 350
Based on NEHRP classification criteria (BSSC, 2003), all the recording stations in the present are classified as 351
either rock site or soil site as given in Column 9, Table 1. Out of 101 recording stations 10 recording stations are 352
classified as rock sites and the rest 91 recording stations are classified as soil sites. 353
5.5 Relationship between GINV H/V results and Vs30 354
The value of are available for 8 recording stations in Tarai region of Uttarakhand and 19 stations in Delhi 355
region from Pandey et al., (2016a) and Pandey et al., (2016b) respectively based on MASW test. These 27 356
recording stations coincides with recording stations where Apeak and fpeak are determined in the present study. 357
Vs30 (obtained from Pandey et al., 2016 a, b), fpeak and Apeak (as per present work) for above 27 recording stations 358
are listed in Table 5. Based on the present findings, relationship between fpeak and Vs30 for the 27 recording 359
stations is proposed as shown in Figure 9a having R2=0.71 as; 360
( )( ) ( ) (16) 361
Similarly, the relationship between Apeak and Vs30 for above 27 recording stations, as obtained in the present 362
study (see Figure 9b) is as follows; 363
( )( ) ( ) (17) 364
A lack of correlation (correlation coefficient =0.47) between and Apeak is observed as shown in Figure 9b, 365
which is also reported in the previous studies like Dutta et al., (2001), (2003) and Hassani et al., (2011). It can 366
be concluded from the proposed correlations (Eqs. 16 and 17) that the value of fpeak increases with increase in 367
Vs30 whereas the value of Apeak decreases with the increase in Vs30. It has to be highlighted here that both the 368
equations [Eqs. 16 and 17] are applicable for sites having fpeak in the range 1.8 to 6 Hz, and Apeak in the range 2 369
to 6.9. 370
Conclusion 371
In the light of on-going seismicity and catastrophic damages witnessed in the past, determination of path 372
attenuation as well as site characterization of EQ recording stations of PESMOS located in the north-west 373
Himalayas and adjoining area is attempted. EQ recorded between 2004 and 2017 are used in the analyses based 374
on two-step inversion. While determining path attenuation, a kink at 105km hypocentral distance is observed in 375
this work. Referring to similar observations from other regions, presence of Moho discontinuity is proposed in 376
the region. This finding can be validated based on detailed study and is beyond the scope of present work 377
objective. Further, based on attenuation curve obtained till 105km hypocentral distance and over wide range of 378
frequencies, ( ) ( ) is obtained for the present study area, clearly indicating that the 379
region is heterogeneous and seismically active. 380
In absence of proper geological information of PESMOS recording stations as highlighted by numerous 381
recent studies, site classification of all the recording stations considered in this study are done based on GINV 382
and HVSR. Based on this analysis, out of 101 recording stations, 10 are found to be located on rock while 91 383
stations are found to be located on soil sites. 1:1 matching for all the recording station from GINV and HVSR 384
further enhances the confidence on present findings. Based on the findings, two empirical correlations for Apeak 385
and fpeak with Vs30, are proposed. 386
Path attenuation and site characteristics are the key factors to be used for developing regional ground 387
motion model. Further, with EQ catalogue known, present findings can be used for detailed seismic hazard 388
assessment of the study area. In addition, identifying recording station whether considered recording stations are 389
located on rock sites or soil sites will help in utilizing ground motion records for scenario based seismic hazard 390
assessment. 391
392
393
Authors Contribution: 394
Harinarayan N H developed code generalized inversion, analyzed the records and all relevant literature review. 395
Kumar Abhishek (AK) highlighted the importance of site characterization for PESMOS recording stations and 396
need for the study. 397
Acknowledgement 398
The authors would like to thank the INSPIRE Faculty program by the Department of Science and Technology 399
(DST), Government of India for the funding project ‗‗Propagation path characterization and determination of in-400
situ slips along different active faults in the Shillong Plateau‘‘ ref. no. DST/INSPIRE/04/2014/002617 [IFA14-401
ENG-104] for providing necessary funding and motivation for the present study. 402
Reference 403
Abdel-Fattah, A. K.: Attenuation of body waves in the crust beneath the vicinity of Cairo Metropolitan area 404
(Egypt) using coda normalization method, Geophys. J. Int., 176(1), 126-134, 2009. 405
Aki, K.: Attenuation of shear-waves in the lithosphere for frequencies from 0.05 to 25 Hz, Phys. Earth Planet. 406
Inter., 21, 50–60, 1980. 407
D‘Alessandro, C., Bonilla, L.F., Boore, D.M., Rovelli, A., and Scotti, O.: Predominant-period site classification 408
for response spectra prediction equations in Italy, Bull. Seismol. Soc. Am., 102(2), 680–695, 2012. 409
Ameri, G., Oth, A., Pilz, M., Bindi, D., Parolai, S., Luzi, L. and Cultrera, G.: Separation of source and site 410
effects by generalized inversiontechnique using the aftershock recordings of the 2009 L‘Aquila earthquake, 411
Bull. of Earthquake Eng., 9(3), 717-739, 2011. 412
Andrews, D. J.: Objective determination of source parameters and similarity of earthquakes of different size, 413
Earthquake source mechanics, 259-267, 1986. 414
Banerjee, S., Kumar, A.: Determination of seismic wave attenuation for the Garhwal Himalayas, India. Geosci 415
Res 2(2):105–126, 2017. 416
Baro, O., Kumar, A., and Zadeh, A.S.: Seismic hazard assessment of the Shillong Plateau, India, Geomatics, 417
Nat. Hazards and Risk, 9:1, 841-861, 2018. 418
Bilham, R., Larson, K., and Freymueller, J.: GPS measurements of present-day convergence across the Nepal 419
Himalaya, Nature, 386(6620), 61, 1997. 420
Bindi, D., Castro, R. R., Franceschina, G., Luzi, L., and Pacor, F.: The 1997–1998 Umbria‐ Marche sequence 421
(central Italy): Source, path, and site effects estimated from strong motion data recorded in the epicentral area, J. 422
Geophys. Res. B: Solid Earth, 109(B4), 2004. 423
Bindi, D., Pacor, F., Luzi, L., Massa, M., and Ameri, G.: The M w 6.3, 2009 L'Aquila earthquake: source, path 424
and site effects from spectral analysis of strong motion data, Geophys. J. Int., 179(3), 1573-1579, 2009. 425
BSSC (2003). Building Seismic Safety Council, NEHRP recommended provisions for seismic regulations for 426
new buildings and other structures. Report FEMA-450 (Provisions), Federal Emergency Management Agency 427
(FEMA), Washington. 428
Boatwright, J., Fletcher, J. B., and Fumal, T. E.: A general inversion scheme for source, site, and propagation 429
characteristics using multiply recorded sets of moderate-sized earthquakes, Bull. Seismol. Soc. Am., 81(5), 430
1754-1782, 1991. 431
Castro, R. R., Anderson, J. G., and Singh, S. K.: Site response, attenuation and source spectra of S waves along 432
the Guerrero, Mexico, subduction zone, Bull. Seismol. Soc. Am., 80(6A), 1481-1503, 1990. 433
Castro, R. R., Fabriol, H., Bour, M., and Le Brun, B.: Attenuation and site effects in the region of Guadeloupe, 434
Lesser Antilles, Bull. Seismol. Soc. Am., 93(2), 612-626, 2003. 435
Castro, R. R., Pacor, F., Sala, A., and Petrungaro, C.: S wave attenuation and site effects in the region of Friuli, 436
Italy, J. Geophys. Res. B: Solid Earth, 101(B10), 22355-22369, 1996. 437
Chung, T. W., and Sato, H.: Attenuation of high-frequency P and S waves in the crust of southeastern South 438
Korea, Bull. Seismol. Soc. Am., 91(6), 1867-1874, 2001. 439
De Lorenzo, S., Del Pezzo, E., and Bianco, F.: Qc, Qβ, Qi and Qs attenuation parameters in the Umbria–Marche 440
(Italy) region, Phys. Earth Planet. Inter. 218, 19-30, 2013. 441
Dutta, U., Biswas, N., Martirosyan, A., Papageorgiou, A., and Kinoshita, S.: Estimation of earthquake source 442
parameters and site response in Anchorage, Alaska from strong-motion network data using generalized 443
inversion method, Phys. Earth Planet. Inter., 137(1-4), 13-29, 2003. 444
Geological Survey of India (GSI) (1998) Geological maps of India. Geological Survey of India, Kolkata 445
Harinarayan, N. H., and Kumar, A.: Site classification of the strong motion stations of Uttarakhand, India, 446
Based on the model horizontal to vertical spectral ratio, Geotechnical Frontiers, 141-149, 2017 a. 447
Harinarayan, N. H., and Kumar, A.: Seismic Site Classification of Recording Stations in Tarai Region of 448
Uttarakhand, from Multiple Approaches, Geotech. Geol. Eng., 1-16, 2017 b. 449
Harinarayan, N. H., and Kumar, A.: Determination of NEHRP Site Class of Seismic Recording Stations in the 450
Northwest Himalayas and Its Adjoining Area Using HVSR Method, Pure Appl. Geophys., 175(1), 89-107, 451
2018. 452
Hartzell, S. H.: Site response estimation from earthquake data, Bull. Seismol. Soc. Am., 82(6), 2308-2327. 453
1992. 454
Hassani, B., Zafarani, H., Farjoodi, J., and Ansari, A.: Estimation of site amplification, attenuation and source 455
spectra of S-waves in the East-Central Iran, Soil Dyn. Earthquake Eng., 31(10), 1397-1413, 2011. 456
IS 1893: Part 1–2016. Indian standard criteria for earthquake resistant design of structures—part 1: General 457 Provisions and Buildings, Bureau of Indian Standards, New Delhi, India. 458
Iwata, T., Irikura, K.: Source parameters of the 1983 Japan Sea earthquake sequence, J. of Physics of the Earth, 459 36(4), 155-184, 1988. 460
Joshi, A., Mohanty, M., Bansal, A. R., Dimri, V. P., and Chadha, R. K.: Use of spectral acceleration data for 461
determination of three-dimensional attenuation structure in the Pithoragarh region of Kumaon Himalaya, J. 462
Seismolog., 14(2), 247-272, 2010. 463
Joshi, A.: Use of acceleration spectra for determining the frequency-dependent attenuation coefficient and 464
source parameters, Bull. Seismol. Soc. Am., 96(6), 2165-2180, 2006. 465
Kim, K. D., Chung, T. W., and Kyung, J. B.: Attenuation of high-frequency P and S waves in the crust of 466
Choongchung provinces, central South Korea, Bull. Seismol. Soc. Am., 94(3), 1070-1078, 2004. 467
Konno, K., and Ohmachi, T.: Ground-motion characteristics estimated from spectral ratio between horizontal 468
and vertical components of microtremor, Bull. Seismol. Soc. Am., 88(1), 228-241, 1998. 469
Kumar, A., Mittal, H., Sachdeva, R., and Kumar, A.: Indian strong motion instrumentation network, Seismol. 470
Res. Lett., 83(1), 59-66, 2012. 471
Kumar, N., Sharma, J., Arora, B. R., and Mukhopadhyay, S.: Seismotectonic model of the Kangra–Chamba 472
sector of northwest Himalaya: Constraints from joint hypocenter determination and focal mechanism, Bull. 473
Seismol. Soc. Am., 99(1), 95-109, 2009. 474
Kvamme, L. B., and Havskov, J.: Q in southern Norway, Bull. Seismol. Soc. Am., 79(5), 1575-1588, 1989. 475
Lermo, J., and Chávez-García, F. J.: Site effect evaluation using spectral ratios with only one station. Bull. 476
Seismol. Soc. Am., 83(5), 1574-1594, 1993. 477
Luzi, L., Puglia, R., Pacor, F., Gallipoli, M.R., Bindi D, and Mucciarelli, M.: Proposal for a soil classification 478
based on parameters alternative or complementary to Vs30, Bull. Seismol. Soc. Am., 9(6), 1877-1898, 2011. 479
Ma‘hood, M., Hamzehloo, H., and Doloei, G. J.: Attenuation of high frequency P and S waves in the crust of the 480
East-Central Iran, Geophys. J. Int., 179(3), 1669-1678, 2009. 481
Mahajan, A.K. and Kumar, S.K.: Macroseismic field observations of January 26th, 2001 Kachchh earthquake 482
and its seismotectonics, J. Asian Earth Sci., 23: 17–23, 2004. 483
Menke, W.: Geophysical data analysis: discrete inverse theory: MATLAB edition (Vol. 45). Academic press, 484
2012. 485
Mukhopadhyay, S., and Kayal, J. R.: Seismic tomography structure of the 1999 Chamoli earthquake source area 486
in the Garhwal Himalaya, Bull. Seismol. Soc. Am., 93(4), 1854-1861, 2003. 487
Mukhopadhyay, S., Sharma, J., Del-Pezzo, E., and Kumar, N.: Study of attenuation mechanism for Garwhal–488
Kumaun Himalayas from analysis of coda of local earthquakes, Phys. Earth Planet. Inter., 180(1-2), 7-15, 2010. 489
Nakamura, Y.: A method for dynamic characteristics estimation of subsurface using microtremor on the ground 490
surface, Railway Technical Research Institute, Quarterly Reports, 30(1), 1989. 491
Nath, S. K., Sengupta, P., and Kayal, J. R.: Determination of S-wave site response in the Garhwal Himalaya 492
from the aftershock sequence of the 1999 Chamoli earthquake, Bull. Seismol. Soc. Am., 92(3), 1072-1081, 493
2002. 494
Negi, S. S., Paul, A., and Joshi, A.: Body wave crustal attenuation characteristics in the Garhwal Himalaya, 495
India, Pure Appl. Geophys., 172(6), 1451-1469, 2015. 496
Oth, A., Bindi, D., Parolai, S., and Di, Giacomo, D.: Spectral Analysis of K-NET and KiK-net Data in Japan, 497
Part II: On Attenuation Characteristics, Source Spectra, and Site Response of Borehole and Surface 498
StationsSpectral Analysis of K-NET and KiK-net Data in Japan, Part II, Bull. Seismol. Soc. Am., 101(2), 667-499
687, 2011. 500
Oth, A., Bindi, D., Parolai, S., and Wenzel, F.: S-wave attenuation characteristics beneath the Vrancea region in 501
Romania: new insights from the inversion of ground-motion spectra, Bull. Seismol. Soc. Am., 98(5), 2482-502
2497, 2008. 503
Oth, A., Parolai, S., Bindi, D., and Wenzel, F.: Source spectra and site response from S waves of intermediate-504
depth Vrancea, Romania, earthquakes, Bull. Seismol. Soc. Am., 99(1), 235-254, 2009. 505
Pandey, B., Jakka, R. S., and Kumar, A.: Influence of local site conditions on strong ground motion 506
characteristics at Tarai region of Uttarakhand, India, Nat. Hazards, 81(2), 1073-1089, 2016 a. 507
Pandey, B., Jakka, R. S., Kumar, A., and Mittal, H.: Site Characterization of Strong‐ Motion Recording Stations 508
of Delhi Using Joint Inversion of Phase Velocity Dispersion and H/V Curve. Bull. Seismol. Soc. Am., 106(3), 509
1254-1266, 2016 b. 510
Parolai, S., Bindi, D., Baumbach, M., Grosser, H., Milkereit, C., Karakisa, S., and Zünbül, S.: Comparison of 511
different site response estimation techniques using aftershocks of the 1999 Izmit earthquake, Bull. Seismol. Soc. 512
Am., 94(3), 1096-1108, 2004. 513
Penrose, R.: A Generalized Inverse for Matrices, Proc. Cambridge Phil. Soc. 51, 406-413, 1955. 514
Philip, G., Suresh, N., and Bhakuni, S.S.: Active tectonics in the northwestern outer Himalaya: evidence of 515
large-magnitude palaeoearthquakes in Pinjaur Dun and the Frontal Himalaya, Current Science, 106, 211–222, 516
2014. 517
Saikia, S., Chopra, S., Baruah, S., Baidya, P. R., and Singh, U.K.: Crustal imaging of the Northwest Himalaya 518
and its foredeep region from teleseismic events, Geomat Nat. Haz. Risk Journal, 7(4), 1265-1286, 2016. 519
Sarkar, I., Pachauri, A.K. and Israil, M.: On the damage caused by the Chamoli earthquake of 29 March, 1999, 520
J. Asian Earth Sci., 19: 129–134, 2001. 521
Sharma, B., Chingtham, P., Sutar, A. K., Chopra, S., and Shukla, H.P.: Frequency dependent attenuation of 522
seismic waves for Delhi and surrounding area, India, Ann. Geophys., 58(2), 0216, 2015. 523
Sharma, J., Chopra, S., and Roy, K. S.: Estimation of source parameters, quality factor (QS), and site 524
characteristics using accelerograms: Uttarakhand Himalaya region, Bull. Seismol. Soc. Am., 104(1), 360-380, 525
2013. 526
Shoji, Y., and Kamiyama, M.: Estimation of local site effects by a generalized inversion scheme using observed 527
records of ‗Small-Titan‘, Soil Dyn. Earthq. Eng., 22(9), 855-864, 2002. 528
Singh, C., Singh, A., Bharathi, V. S., Bansal, A. R., and Chadha, R. K.: Frequency-dependent body wave 529
attenuation characteristics in the Kumaun Himalaya, Tectonophysics, 524, 37-42, 2012. 530
Srivastava, H.N., Bansal, B.K., and Verma, M.: Largest earthquake in Himalaya: An appraisal, J. Geol. Soc. 531
India, 82, 15–22, 2013. 532
Tripathi, J. N., Singh, P., and Sharma, M. L.: Attenuation of high-frequency P and S waves in Garhwal 533
Himalaya, India, Tectonophysics, 636, 216-227, 2014. 534
Valdiya, K.S.: Aspects of Tectonics, Focus on south-central Asia, Tata McGraw-Hill Publishing Company Ltd., 535
New Delhi, 319, 1984. 536
Yaghmaei-Sabegh, S.: Characteristics of near-source ground motions from the 2012 Varzaghan-Ahar double 537
Earthquakes, Northwest of Iran, Nat. Hazards, 70, 1077–1097, 2014. 538
Yoshimoto, K., Sato, H., and Ohtake, M.: Frequency-dependent attenuation of P and S waves in the Kanto area, 539
Japan, based on the coda-normalization method, Geophys. J. Int., 114(1), 165-174, 1993. 540
541
Appendix A 542
In the matrix form, following the notations of Menke (1989) and incorporating the weighting factors 543
and , Eq. (1) can be written in accordance with Castro et al., (1990) as: 544
(A) (X) (b) 545
(A1) 546
The hypocentral distances of the data set is discretized into number of bins of equal widths and the value of 547
( ) is computed at each bin. The width of the bins are selected such that there is almost equal number of 548
data points in every bin. Further, ( ) versus hypocentral distance curves at each of the selected 549
frequencies are computed solving Eq. (2) in a least square sense, using singular value decomposition method 550
(Menke, 1989). 551
552
553
1 0 0 . …. 1 0 0 . ….
ln A1
ln U11
0 1 0 . …. 1 0 0 . ….
.
.
. . . . …. . . . . ….
.
.
1 0 0 . …. 0 1 0 . ….
. = ln Uij
. . . . . 0 1 0 . ….
ln A10
.
. . . . . . . . . . .
ln M1 0 . . . . . . . . . .
w1 0 0 . …. . . . . ….
.
0
- w 2/2 w 2 - w 2/2 . …. . . . . ….
.
0
0 - w 2/2 w 2 - w2/2 …. . . . . ….
.
0
. . . . …. . . . . ….
ln M N
.
FIGURES 554
555
Figure 1: Map of the region under study with EQs (stars), recording stations (triangles), and paths (solid-lines). 556
557
Figure 2: Distribution of hypocentral distances in the data set. 558
559
Figure 3: S wave spectral attenuation versus hypocentral distance. Note that ln A(f,R0) at reference distance is zero. 560
29
18
2021
2223
21
1817
13
9
2
01
3
0
5
10
15
20
25
30
35
Num
ber
of
reco
rds
Hypocentral distance (km)15 25 35 45 55 65 75 85 95 105 115 125 135 145 155
-3
-2.5
-2
-1.5
-1
-0.5
0
15 30 45 60 75 90 105
ln A
(f,R
)
R (km)
0.5Hz 1Hz
1.75Hz 2.5Hz
3.125Hz 4.5Hz
5.5Hz 6.25Hz
8Hz 10Hz
11.76Hz 12.5Hz
13.33Hz 14.28Hz
15Hz
561
Figure 4: Frequency dependence of the quality factor Q for hypocentral distances between 15 km to 105 km 562
563
Figure 5: Comparison of QS values of North West Himalaya with those obtained from parts of North West Himalaya 564 and Delhi region. The compared relations for Qs versus frequency are as follows: Garhwal-Kumouan Himalaya: 565
(Mukhopadhyay et al., 2010); Kinnaur Himalaya: (Kumar et al., 2014) ; Garhwal 566 Himalaya: (Negi et al., 2015); Delhi and NCR region: (Sharma et al., 2015). 567
568
Qs= 105 f 0.94
R² = 0.97
0
200
400
600
800
1000
1200
1400
1600
0 5 10 15
Q
frequency (Hz)
0
400
800
1200
1600
2000
0.5 5.5 10.5 15.5
Qs
frequency (Hz)
Garhwal-KumouanHimalaya(Mukhopadhyay etal., 2010)
Kinnaur Himalaya(Kumar et al., 2014)
Garhwal Himalaya(Negi et al., 2015)
Delhi and NCRregion (Sharma et al.,2015)
North WestHimalaya-This Study
569
Figure 6: Comparison of QS values of this study with regions of different tectonic settings of the world. 570
571
572
1
1001
2001
3001
0.5 5.5 10.5 15.5
Qs
frequency(Hz)
North West Himalaya-This study
Kato region-Japan(Yoshimoto et al.,1995)East central Iran
(Ma‘hood et al., 2009)
North East India(Padhy and Subhadra2010)Mainland Gujarathregion (Chopra et al.,2011)Egypt (Abdel 2009)
Koyna Region India(Sharma at al., 2007)
Umbria Marche-Italy(Lorenzo et al., 2013)
South Central Alaska(Dutta et al., 2004)
0
0.04
0.08
0.12
0.15 1.5 15
SA
(g)
Frequency (Hz)
(a) ANC
GINV H
GINV V
573
574
0
0.3
0.6
0.15 1.5 15
SA
(g)
Frequency (Hz)
(b) IGN
GINV H
GINV V
0
0.2
0.4
0.15 1.5 15
SA
(g)
Frequency (Hz)
(c) JAMI
GINV H
GINV V
575
576
0
0.03
0.06
0.09
0.15 1.5 15
SA
(g)
Frequency (Hz)
(d) BAR
GINV H
GINV V
0
0.05
0.1
0.15
0.15 1.5 15
SA
(g)
Frequency (Hz)
(e) GHA
GINV H
GINV V
577 578
Figure 7 (a-f). Site amplitude curves obtained using GINV for horizontal component and vertical component 579
580
0
0.1
0.2
0.3
0.15 1.5 15
SA
(g)
Frequency (Hz)
(f) SND
GINV H
GINV V
0
1
2
3
4
5
6
7
0.15 1.5 15
H/V
Frequency (Hz)
(a) ALIP
HVSR
GINV
581
582
0
0.5
1
1.5
2
2.5
3
3.5
4
0.15 1.5 15
H/V
Frequency (Hz)
(b) IGI
HVSR
GINV
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
5
0.15 1.5 15
H/V
Frequency (Hz)
(c) IIT
HVSR
GINV
583
584
0
0.5
1
1.5
2
2.5
3
3.5
0.15 1.5 15
H/V
Frequency (Hz)
(d) PALHVSR
GINV
0
1
2
3
4
5
6
0.15 1.5 15
H/V
Frequency (Hz)
(e) BAL
HVSR
GINV
585
586
0
1
2
3
4
5
6
0.15 1.5 15
H/V
Frequency (Hz)
(f) RAM
HVSR
GINV
0
1
2
3
4
5
6
0.15 1.5 15
H/V
Frequency (Hz)
(g) SAL
HVSR
GINV
587
588
Figure 8 (a-i): Horizontal to vertical ratio curve obtained using GINV and HVSR method. 589
590
591
592
593
594
595
0
2
4
6
8
10
12
14
16
18
0.15 1.5 15
H/V
Frequency (Hz)
(h) MUK
GINV
HVSR
0
1
2
3
4
5
6
7
8
9
0.15 1.5 15
H/V
Frequency (Hz)
(i) PTI
HVSR
GINV
596
597
Figure 9: Vs30 as a function of fpeak (a) and Apeak (b) of GINV method for recording stations at Delhi and Tarai region 598 of Uttarakhand. 599
600
601
602
603 604
605
log (Vs30 )= 0.48 log (fpeak)+ 2.3
R² = 0.71
2.3
2.4
2.5
2.6
2.7
2.8
0 0.2 0.4 0.6 0.8
log (
Vs3
0)
(m/s
)
log (fpeak)
(a)
log (Vs30) = -0.74 log (Apeak) + 2.93
R² = 0.482.2
2.3
2.4
2.5
2.6
2.7
2.8
0.2 0.4 0.6 0.8
log (
Vs3
0)
(m/s
)
log (Apeak)
(b)
TABLES 606
Table 1: Detail of strong motion recording stations. 607
Si.
no
Station
Code
GINV HVSR
R
* or
S#
Si.
no
Station
Code
GINV HVSR
R
* or
S#
Lat.(°
) (N)
Lon. (°)
(E) fpeak
Apea
k fpeak
Apea
k
Lat.(°
) (N)
Lon. (°)
(E)
fpea
k
Apea
k
fpea
k
Apea
k
-1 -2 -3 -4 -5 -6 -7 -8 -9 -1 -2 -3 -4 -5 -6 -7 -8 -9
Himachal Pradesh
Uttarakhand
1 AMB 31.7 76.1 1.7 4.3 1.2 9.3 S 1 ALM 29.6 79.7 2.1 3.0 2.8 4.4 S
2 BHA 31.6 77.9 4.5 4.0 4.1 4.4 S 2 BAG 29.8 79.8 1.5 4.7 1.5 5.2 S
3 CHM 30.4 79.3 1.4 5.4 1.5 7.5 S 3 BAR 30.8 78.2 3.0 4.5 2.8 7.0 S
4 DEH 31.9 76.2 6.8 3.5 10 5.4 R 4 CHM 32.6 76.1 3.6 2.4 2.0 2.9 S
5 DHH 32.2 76.3 2.7 4.5 2.7 4.9 S 5 CHP 29.3 80.1 5.4 5.2 5.6 6.5 S
6 HAM 31.7 76.5 2.9 3.3 3.1 6.6 S 6 CKR 30.7 77.9 2.1 3.8 2.0 4.5 S
7 JUB 31.1 77.7 5.8 3.3 5.6 4.9 S 7
CMB
B 30.0 79.5 8.3 2.7 8.3 6.3 R
8 KLG 32.6 77.0 8.0 1.8 8.3 2.2 R 8 DHA 29.8 80.5 3.1 3.3 2.7 5.5 S
9 KUL 32.0 77.1 3.3 2.3 3.1 3.0 S 9 DNL 30.4 78.2 2.8 3.3 2.0 7.1 S
10 MAN 31.7 76.9 2.5 3.4 2.3 5.3 S 10 DUN 30.3 78.0 2.9 5.8 3.1 7.1 S
11 RAM 31.4 77.6 2.8 4.2 2.7 5.6 S 11 GAR 30.1 79.3 2.4 3.4 2.3 4.5 S
12 SAL 32.7 76.1 2.7 3.0 2.7 5.8 S 12 GHA 30.4 78.7 5.2 2.3 4.5 5.5 S
13 SND 31.5 76.9 5.0 3.5 5.0 4.2 S 13 GLTR 30.3 79.1 2.9 3.5 2.8 6.1 S
14 SOL 30.9 77.1 0.5 2.6 0.6 4.6 S 14 JSH 30.5 79.6 1.4 2.2 1.5 3.0 S
15 UNA 31.5 76.3 1.5 3.9 1.8 6.0 S 15 KAP 29.9 79.9 3.7 6.4 3.3 9.2 S
16 KJK 30.9 76.9 0.7 6.4 0.4 12.0 S 16 KHA 28.9 80.0 1.3 4.2 2.0 8.0 S
17 PLM 32.1 76.5 1.7 1.8 1.6 2.2 S 17 KKHR 30.2 78.9 9.5 3.5 9.5 9.3 R
Punjab
18 KOT 29.7 78.5 0.7 2.4 0.7 3.3 S
1 ANS 31.2 76.5 0.9 4.9 0.9 17.0 S 19 KSK 29.2 79.0 3.1 3.8 3.2 9.9 S
2 ASR 31.6 74.9 1.0 2.9 1.0 8.0 S 20 KSL 30.9 77.0 3.0 3.9 2.1 19.4 S
3 GSK 31.2 76.1 0.8 2.3 0.8 3.8 S 21 LANG 30.3 79.3 7.7 2.8 7.9 5.7 R
4 JAL 31.3 75.6 2.9 1.5 2.9 1.7 S 22 LAN 29.8 78.7 1.4 2.8 1.4 6.1 S
5 KAT 31.4 75.4 2.0 3.1 3.0 3.4 S 23 MUN 30.1 80.2 4.3 2.8 7.0 3.6 S
6 MOG 30.8 75.2 2.2 2.3 2.2 3.9 S 24 PAU 30.2 78.8 5.9 2.1 3.1 3.7 S
7 MUK 31.9 75.6 1.4 4.7 1.4 15.9 S 25 PTH 29.6 80.2 8.0 2.7 4.6 3.1 R
8 NAW 31.1 76.1 1.4 3.6 1.4 3.5 S 26 PTI 29.4 79.9 4.0 6.6 3.6 8.0 S
9 NKD 31.1 75.5 1.2 2.2 1.3 3.4 S 27 RIS 30.1 78.3 3.8 3.0 3.4 6.4 S
10 PHG 31.2 75.8 2.7 2.7 2.7 5.2 S 28 ROO 29.9 77.9 1.2 4.4 1.3 5.2 S
11 TAR 31.4 74.9 2.6 2.7 2.7 4.6 S 29 RUD 30.3 79.0 1.3 2.8 1.5 4.2 S
Delhi
30 SMLI 30.2 79.3 9.1 3.3 8.7 6.1 R
1 ARI 26.1 77.5 2.7 3.2 2.5 7.0 S 31 TAN 29.1 80.1 5.4 3.4 5.0 6.3 S
2 IGN 28.5 77.2 3.6 2.6 4.5 3.9 S 32 THE 30.4 78.4 1.6 2.8 1.5 3.6 S
3 JNU 28.5 77.2 9.0 2.1 8.7 3.5 R 33 UDH 29.0 79.4 2.7 6.9 2.2 10.1 S
4 DJB 28.7 77.2 2.2 3.2 10 4.5 S 34 UTK 30.7 78.4 2.4 2.9 2.3 4.4 S
5 NDI 28.7 77.2 6.8 2.5 7.2 3.7 R 35 VIK 30.5 77.8 2.3 3.8 2.3 10.4 S
6 IMD 28.7 77.2 6.0 2.0 6.3 2.9 S 36 GDRI 30.2 78.7 6.0 3.4 5.1 4.8 S
7 NTPC 28.5 77.3 2.8 3.6 2.8 5.4 S 37 TLWR 30.3 79.0 1.1 2.1 1.0 4.6 S
8 ANC 28.5 77.3 4.6 3.2 4.5 4.5 S 38 UKMB 30.3 79.1 1.0 2.9 1.4 10.0 S
9 JAMI 28.6 77.3 4.7 3.2 4.5 7.3 S 39 ADIB 30.2 79.2 6.4 12.7 6.3 17.8 R
10 LDR 28.6 77.2 0.7 4.3 0.9 7.0 S 40 NUTY 30.2 79.2 4.8 2.4 4.7 4.3 S
11 VCD 28.6 77.2 4.6 2.7 4.6 3.6 S 41 KHIB 30.2 78.8 7.7 3.3 8.0 8.7 R
12 IIT 28.6 77.3 4.3 2.9 4.5 4.3 S 42 STRK 30.3 79.0 4.7 2.8 4.7 5.8 S
13 NSIT 28.6 77.0 2.4 2.5 2.3 3.9 S 43 NANP 30.3 79.3 3.9 3.5 3.8 9.1 S
14 RGD 28.7 77.1 2.3 2.1 2.9 3.8 S
Haryana
15 GGI 28.7 77.2 15 5.3 15 8.4 R 1 PAL 28.1 77.3 2.8 2.7 2.9 3.4 S
16 DLU 28.7 77.2 1.8 3.2 1.9 3.7 S 2 JAFR 28.6 76.9 6.0 2.0 7.1 2.6 S
17 DCE 28.8 77.1 3.8 3.1 4.7 4.2 S 3 GUR 28.4 77.0 1.0 4.1 1.0 5.2 S
18 IGI 28.6 77.1 2.2 2.4 2.2 3.8 S 4 REW 28.2 76.6 2.5 2.1 2.5 3.7 S
19 ZAKI 28.6 77.2 3.9 3.5 3.9 8.4 S 5 SON 29.0 77.0 1.0 3.5 2.8 4.0 S
20 ALIP 28.8 77.1 2.3 3.2 2.5 6.9 S 6 ROH 28.6 77.2 1.4 3.1 2.0 4.6 S
21 ROI 28.6 77.2 1.4 3.1 2.0 4.6 S 7 CRRI 29.0 77.1 4.3 3.5 4.4 9.3 S
R* Rock site
8 BAL 28.3 77.3 1.5 3.0 1.4 5.8 S
S# Soil site
9 KAI 29.8 76.4 1.2 3.0 1.2 6.5 S
608
Table 2: Details of earthquake events (from PESMOS) considered for estimation of site parameters in 609 this work. 610
Eve
nt
No.
dd/mm/yy
yy
Lat
.
Lon
g.
Dept
h
Magnitu
de
Eve
nt
No.
dd/mm/yy
yy
Lat
.
Lon
g.
Dept
h
Magnitu
de
-1 -6 -2 -3 -4 -5
-1 -6 -2 -3 -4 -5
1
14-12-
2005
30.
9 79.3 25.7 5.2
44
24-09-
2011
30.
9 78.3 10.0 3.0
2
07-05-
2006
28.
7 76.6 20.2 4.1
45
26-10-
2011
31.
5 76.8 5.0 3.5
3
29-11-
2006
27.
6 76.7 13.0 3.9
46
16-01-
2012
29.
7 78.9 10.0 3.6
4
10-12-
2006
31.
5 76.7 33.0 3.5
47
12-03-
2012
28.
9 77.3 5.0 3.5
5
22-07-
2007
29.
9 77.9 33.0 5.0
48
26-02-
2012
29.
6 80.8 10.0 4.3
6
25-11-
2007
28.
6 77.0 20.3 4.3
49
27-03-
2012
26.
1 87.8 12.0 3.5
7
04-10-
2007
32.
5 76.0 10.0 3.8
50
05-03-
2012
28.
7 76.6 14.0 4.9
8
18-10-
2007
28.
3 77.6 5.6 3.6
51
28-07-
2012
29.
7 80.7 10.0 4.5
9
19-08-
2008
30.
1 80.1 15.0 4.3
52
23-08-
2012
28.
4 82.7 5.0 5.0
10
19-10-
2008
29.
1 76.9 7.0 3.2
53
02-10-
2012
32.
4 76.4 10.0 4.9
11
21-10-
2008
31.
5 77.3 10.0 4.5
54
03-10-
2012
32.
4 76.3 10.0 3.6
12
31-01-
2009
32.
5 75.9 10.0 3.7
55
06-11-
2012
32.
3 76.2 5.0 4.1
13
09-01-
2009
31.
7 78.3 16.0 3.8
56
11-11-
2012
29.
3 80.1 5.0 5.0
14
25-02-
2009
30.
6 79.3 10.0 3.7
57
15-11-
2012
30.
2 80.1 5.0 3.0
15
18-03-
2009
30.
9 78.2 10.0 3.3
58
27-11-
2012
30.
9 78.4 12.0 4.8
16
04-09-
2008
30.
1 80.4 10.0 5.1
59
19-12-
2012
28.
6 76.8 10.0 2.9
17 01-05- 29. 80.1 10.0 4.6
60 02-01- 29. 81.1 10.0 4.8
2009 9 2013 4
18
15-05-
2009
30.
5 79.3 15.0 4.1
61
09-01-
2013
29.
8 81.7 5.0 5.0
19
17-07-
2009
32.
3 76.1 39.3 3.7
62
10-01-
2013
30.
1 80.4 5.0 3.2
20
27-08-
2009
30.
0 80.0 14.0 3.9
63
29-01-
2013
30.
0 81.6 7.0 4.0
21
21-09-
2009
30.
9 79.1 13.0 4.7
64
11-02-
2013
31.
0 78.4 5.0 4.3
22
03-10-
2009
30.
0 79.9 15.0 4.3
65
17-02-
2013
30.
9 78.4 10.0 3.2
23
06-12-
2009
35.
8 77.3 60.0 5.3
66
01-05-
2013
33.
1 75.8 15.0 5.8
24
11-01-
2010
29.
7 80.0 15.0 3.9
67
05-09-
2013
30.
9 78.5 11.0 3.5
25
22-02-
2010
30.
0 80.1 2.0 4.7
68
11-11-
2013
28.
5 77.2 10.0 3.1
26
24-02-
2010
28.
6 76.9 17.0 2.5
69
11-11-
2013
28.
4 77.2 11.0 2.8
27
14-03-
2010
31.
7 76.1 29.0 4.6
70
11-11-
2013
28.
4 77.2 12.0 2.5
28
03-05-
2010
30.
4 78.4 8.0 3.5
71
11-11-
2013
28.
4 77.2 13.0 3.1
29
28-05-
2010
31.
2 77.9 43.0 4.8
72
16-04-
2013
28.
0 62.1 16.0 7.8
30
31-05-
2010
30.
0 79.8 10.0 3.6
73
04-06-
2013
32.
7 76.7 18.0 4.8
31
06-07-
2010
29.
8 80.4 10.0 5.1
74
05-06-
2013
32.
8 76.3 10.0 4.5
32
10-07-
2010
29.
9 79.6 10.0 4.1
75
09-07-
2013
32.
9 78.4 10.0 5.1
33
26-01-
2011
29.
0 77.2 10.0 3.2
76
13-07-
2013
32.
2 76.3 10.0 10.0
34
14-03-
2011
30.
5 79.1 8.0 3.3
77
15-07-
2013
32.
6 76.7 30.0 4.4
35
18-02-
2011
28.
6 77.3 5.0 2.3
78
02-08-
2013
33.
5 75.5 20.0 5.4
36
09-02-
2011
30.
9 78.2 10.0 5.0
79
29-08-
2013
31.
4 76.1 10.0 4.7
37
04-04-
2011
29.
6 80.8 10.0 5.7
80
20-10-
2013
35.
8 77.5 80.0 5.5
38
15-06-
2011
30.
6 80.1 10.0 3.6
81
25-12-
2013
31.
2 78.3 10.0 4.0
39
20-06-
2011
30.
5 79.4 12.0 4.6
82
17-06-
2014
32.
2 76.1 10.0 4.1
40
23-06-
2011
30.
0 80.5 5.0 3.2
83
21-08-
2014
32.
3 76.5 10.0 5.0
41
28-07-
2011
33.
3 76.0 21.0 4.4
84
29-11-
2015
30.
6 79.6 15.0 4.0
42
07-09-
2011
28.
6 77.0 8.0 4.2
85
25-09-
2016
30.
0 79.5 11.0 3.7
43
21-09-
2011
30.
9 78.3 10.0 3.1
86
01-12-
2016
30.
6 79.6 19.0 4.0
611
Table 3: List of Earthquakes and the corresponding stations considered for the estimation of path 612 parameter. 613
Earthquake Event Stations
25-11- 2007 HGR, NDI, CRRI, PAL, REW, NDI , CRRI, LDR, JAFR, IIT
19-08-2008 CHP, PTH, KAP, MUN
04-09- 2008 MUN, CHP, PTH, DHA , KAP, GHA, JSH
01-05-2009 MUN, BAG, KAP, GAR, CHM
17-07-2009 DHA, KLG
27-08-2009 KAP, BAG, MUN
03-08-2009 KAP, CHP, BAG
11-01-2010 PTH, CHP, DHA
22-02-2010 KAP, BAG, DHA, ROO, UDH
24-02-2010 RGD, IGN, ROH, DJB, CHP, ANC, JAMI, GGI, DLU, DCE
14-03-2010 DEH, JUB, SND, BHA, HAM, GAR, JAL, KAP, AMB, ROO
03-04-2010 THE, BAR, DNL, ROO
28-05-2010 JUB, BAR, ROO, UNA
06-06-2010 MUN, CHP
10-07-2010 BAG, KAP, GAR, ROO
18-02-2011 DJB, ANC
09-02-2011 UTK, SND, KUL, CKR
04-04-2011
JSH, CHP, PTI, PTH, ALM, DDH, BAG, DHA, GAR, MUN, RUD, THE, CHM,
BAR, SND, KOT, DNL, LDR, ROO, TAN, KHA, UDH, KSH, DUD
12-03-2012 GGI, ANC, DLU, DCE
27-03-2012 ARI, ANC
05-03-2012
JAFR, JNU, DJB, IMD, PLW, GUR, NOI, NTPC, ANC, IIT, NSIT, ZAKI, ROO,
RGD, GGI, DLU, DCE, ALIP, SON, BAR, KAI, NKD,
02-10-2012 CHA, RAM
11-11-2012 CHA, CHP, PTH
27-11-2012 UTK, THE, DNL, CKR
02-01-2013 CHP, PTI, PTH
09-01-2013 CHP, PTI, PTH, TAN
11-02-2013 UTK, ROO
11-11-2013 (19:11:18) NTPC, IGN, JNU, DJB, IMD, VCD, IGI, RGD, GGI, DLU, DCE, ALIP
11-11-2013 (22,10,42) IGN, JNU, DJB, VCD, RGD, DCE, ALIP
11-11-2013 (20:11:30) IGN, JNU, DJB, VCD, RGD, GGI, DLU, DCE
29-08- 2013 GSK, RAM, ROO, NKD, ANS, KAT
25-09-2016 UKMB, CMBB, GDRI, DURD
614
Table 4: Resulting parameters of eq. 7. 615
f (Hz) m
( )
( )⁄
(1) (2) (3)
0.50 0.0095 51.65
1.00 0.0101 97.15
1.75 0.0115 149.32
2.50 0.0096 255.53
3.12 0.0089 344.54
3.57 0.0093 376.67
4.50 0.0098 450.57
5.00 0.0074 663.01
5.50 0.0089 606.39
6.25 0.0084 730.09
7.14 0.011 636.92
8.00 0.013 603.84
10.00 0.0172 570.49
11.76 0.0124 930.60
12.50 0.0102 1202.51
13.33 0.0098 1334.70
14.28 0.0115 1218.46
15.00 0.0108 1362.85
Table 5: fpeak, Apeak and Vs30 values for 27 stations located in Terai region of Uttarakhand and Delhi 616 region. 617
GINV
Station
Code fpeak,(Hz) Apeak, Vs30 (m/s)
IGN 3.6 2.6 493*
JNU 6.5 2.05 565*
DJB 5.22 3.2 543*
NDI 6.8 2.5 493*
IMD 6 2 543*
NTPC 2.8 3.6 345*
ANC 4.6 3.2 564*
JAMI 4.7 3.15 346*
LDR 0.7 4.32 270*
VCD 5.6 2.7 550*
IIT 4.3 2.94 332*
NSIT 2.4 2.5 391*
RGD 2.3 2.07 346*
DLU 1.81 3.2 323*
DCE 3.78 3.1 328*
IGI 2.2 2.4 360*
ZAKI 3 3.49 337*
ROI 2.3 3.24 303*
ALIP 1.4 3.09 338*
DUN 2.9 5.8 289**
KHA 1.3 4.2 218**
KSK 3.13 3.75 208**
RIS 3.8 3 331**
ROO 1.16 4.35 218**
TAN 5.4 3.4 434**
UDH 2.74 6.9 198**
VIK 2.29 3.8 424**
**Pandey et al., 2016a; * Pandey et al., 2016b 618
619
620
621
622
623
624
625
626
627
List of Figures 628
1. Figure 1: Map of the region under study with EQs (stars), recording stations (triangles), and paths (solid-629 lines). 630
2. Figure 2: Distribution of hypocentral distances in the data set. 631
3. Figure 3: S wave spectral attenuation versus hypocentral distance. Note that Log A(f,R0) at reference distance 632 is zero. 633
4. Figure 4: Frequency dependence of the quality factor Q for hypocentral distances between 15km to 105km 634
5. Figure 5: Comparison of QS values of North West Himalaya with those obtained from parts of North West 635 Himalaya and Delhi region. The compared relations for Qs versus frequency are as follows: Garhwal-Kumouan 636 Himalaya: (Mukhopadhyay et al., 2010); Kinnaur Himalaya: (Kumar et al., 637 2014) ; Garhwal Himalaya: (Negi et al., 2015); Delhi and NCR region: 638 (Sharma et al., 2015). 639
6. Figure 6: Comparison of QS values of this study with regions of different tectonic settings of the world. 640
7. Figure 7: Site amplitude curves obtained using GINV for horizontal component and vertical component 641
8. Figure 8: Horizontal to vertical ratio curve obtained using GINV and HVSR method 642
9. Figure 9: Vs30 as a function of fpeak (a) and Apeak (b) of GINV method for recording stations at Delhi and 643 Tarai region of Uttarakhand. 644
List of Tables 645
Table 1: Detail of strong motion recording stations. 646
Table 2: Details of earthquakes considered for estimation of site parameters in this work 647
Table 3: List of Earthquakes and the corresponding stations considered for the estimation of path parameter. 648
Table 4: Resulting parameters of Eq. (7). 649
Table 5: fpeak, Apeak and Vs30 values for 27 stations located in Terai region of Uttarakhand and Delhi region. 650
651